Le Monde Rich
Went to Richmond yesterday for a game of Go with Daniel. It was probably the most enjoyable game we've played to date. When we started, Dan had been utterly addicted to TurboGo, and regularly trounced me. We broke off from playing for quite some time, and when we finally did play, I started trouncing him. In fact in our last game I very nearly beat him by 160 points, but cocked up in the endgame, letting a dead group live by playing diagonally into his 2x2 territory. Are you with me so far?
When Dan was beating me, it always seemed like there were certain turning points in the game that tipped the balance hugely in his favour - small errors with massive consequences. Yesterday's game felt a little like that, except that it tipped in my favour instead.
I came away from the game more determined to keep playing against humans as much as possible. I have Many Faces Of Go, but tend to try and outpace the machine's playing speed, meaning I play quickly and badly. I should learn to sit on my hands. I went back to a really strange game I played on KGS this morning, running through it looking where I'd screwed up, which was an eye-opener. I took black, and managed to force white into a massive dumpling shape in the middle of the board, with enough of my stones around it to cut out a serious amount of territory. However, white managed a few substantial invasions, and I lost my lead.
Going back over it, I realised I'd played better than I'd thought, despite the 50 points I lost by. I find it strange how easy it is to detach myself from the game when going through it like that. The KGS client software is a dream for this sort of work, too, offering a tree diagram system for exploring variations, a generous comment box, and the usual array of stone marking tools. What I'd really like to be able to do is rig up Many Faces to the KGS client, but there are obvious reasons why that wouldn't be easy. Regardless of that, going over old games really makes it obvious where your mistakes have been, and where you should investigate and study.
Hoping to go back to Richmond on Wednesday, work allowing. I finish my current placement on Tuesday, but would have to take more work if I was offered it.
When Dan was beating me, it always seemed like there were certain turning points in the game that tipped the balance hugely in his favour - small errors with massive consequences. Yesterday's game felt a little like that, except that it tipped in my favour instead.
I came away from the game more determined to keep playing against humans as much as possible. I have Many Faces Of Go, but tend to try and outpace the machine's playing speed, meaning I play quickly and badly. I should learn to sit on my hands. I went back to a really strange game I played on KGS this morning, running through it looking where I'd screwed up, which was an eye-opener. I took black, and managed to force white into a massive dumpling shape in the middle of the board, with enough of my stones around it to cut out a serious amount of territory. However, white managed a few substantial invasions, and I lost my lead.
Going back over it, I realised I'd played better than I'd thought, despite the 50 points I lost by. I find it strange how easy it is to detach myself from the game when going through it like that. The KGS client software is a dream for this sort of work, too, offering a tree diagram system for exploring variations, a generous comment box, and the usual array of stone marking tools. What I'd really like to be able to do is rig up Many Faces to the KGS client, but there are obvious reasons why that wouldn't be easy. Regardless of that, going over old games really makes it obvious where your mistakes have been, and where you should investigate and study.
Hoping to go back to Richmond on Wednesday, work allowing. I finish my current placement on Tuesday, but would have to take more work if I was offered it.
posted by Hamilton's Brain | 9:17 pm
0 Comments:
Post a Comment
<< Home